Management of Idiosyncratic Hepatotoxicity

MANAGEMENT OF IDIOSYNCRATIC HEPATOTOXICITY

Early detection and withdrawal of the causative drug is the single most important step in the management of adverse hepatic reaction. Cases of serious and often fatal hepatotoxicity due to isoniazid, halothane, valproate, nitrofurantoin and perhexiline are often linked to continuation or resumption of the drug following symptoms that could have been attributable to drug-induced liver reaction (Farrell, 1994; Lo et al., 1998; Moulding, 1999). The Seattle-King County Public Health Department used a protocol to monitor isoniazid therapy, which included advising the patient at each visit to stop the medication and call the clinic if symptoms of hepatotoxicity occurred. With care-ful monitoring, the rate of hepatotoxicity in 11 141 patients was much lower (0.1%–0.15%) than previ-ously reported (1%), and there were no deaths (Nolan, Goldberg and Buskin, 1999). Prompt withdrawal of the drug is also important because the long-term prog-nosis may be worse if the responsible agent is contin-ued. In a retrospective study, one-third of patients with drug-induced liver disease had persistently abnormal liver tests (liver enzymes and/or imaging) at median follow-up of 5 years, and the detection of fibrosis in the liver biopsy and continued drug intake after the initial liver injury predicted adverse outcome (Aithal and Day, 1999).

Management of acute hepatic failure secondary to idiosyncratic hepatic reaction is similar to that of viral hepatitis. The overall mortality of drug-induced hepatic failure (excluding paracetamol over-dose) appears to be higher than that of viral hepatitis. Despite the availability of liver transplantation, 13% of those who develop jaundice due to severe hepa-totoxicity die, and in patients with halothane-induced liver injury, the mortality rate of 40% have been reported (Bjornsson and Olsson, 2005). Corticosteroid treatment has not been shown to be beneficial in the management of drug-induced hepatitis. There is no clear evidence that ursodeoxycholic acid therapy changes outcome in chronic cholestasis.

PREVENTION

Experience gained by wide clinical usage of a drug following marketing may assist in recognising indi-vidual risk factors and better definition of safe dosage. Strategies of avoiding the prescription in ‘at-risk situ-ations’ and safer dosage regimes have reduced adverse hepatic reactions due to several drugs. Some such examples include the avoidance of reuse of halothane within 3 months, parenteral administration of large doses of tetracycline as well as its use in pregnancy and renal disease, aspirin in children and valproic acid in combination therapy in children under the age of 3 years (Farrell, 1994; Neuberger, 1998). The incidence of hepatic fibrosis with weekly low-dose methotrexate regimes is much lower than that reported with daily dose regimes (Boffa et al., 1995; Aithal et al., 2004a).

When a new drug is recognised to be associated with significant hepatic ADR, it has become common prac-tice to recommend regular monitoring of liver enzymes for the early detection of liver injury so that drug can be withdrawn before serious hepatotoxic reaction occurs. Although this is logical, the level of enzyme elevation at which the risk of serious, progressive hepatotoxicity is significant and yet the injury is completely reversible on the withdrawal of medication is still uncertain (Kaplowitz, 2005). In addition, compliance with such recommendation remains low (Gaham et al., 2001).

Of even greater importance in the determination of individual risk is the inherited factors that affect the kinetics and dynamics of numerous drugs. Suscepti-bility to hepatic drug reaction depends principally on genetic factors that determine the metabolism, as well as the biochemical and immunological responses, to the metabolites. A major difference between genetic and environmental variation is that an inherited trait has to be tested for only once in a lifetime, whereas environmental effects change continuously. In the future, the discovery of pharmacogenetic traits will change with new technologies based on genomics. Rapid sequencing and single-nucleotide polymor-phisms (SNPs) will play a major role in the linking of sequence variations with heritable phenotypes of drug response (Meyer, 2000). In fact, pharmacoge-netics technology may enable a significantly better post-marketing surveillance system. In this proposed concept (Roses, 2000), hundreds of thousands of patients who receive the drug would have blood taken and stored in an approved location. As rare, serious adverse events are documented, DNA from patients who experienced the ADR could be compared with that from controls, who did not have adverse reaction while on the drug. This would enable ‘genetic finger-prints’ (SNP profiles) of the subjects susceptible to the adverse event to be determined. These adverse event profiles would be combined with efficacy profiles to produce a comprehensive medicine response profile. This would allow the selection of patients for both efficacy and lower complications of drug therapy.